Exploratory Data Analysis of Streamable Movies

Import Libraries

import pandas as pd
pd.options.mode.chained_assignment = None
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt
%matplotlib inline
import os
import plotly.express as px

Read in the File

raw_movies_df = pd.read_csv('../csv/MoviesOnStreamingPlatforms_updated.csv', index_col=1)

raw_movies_df.drop(columns='Unnamed: 0', inplace=True)

raw_movies_df.head()

	Title	Year	Age	IMDb	Rotten Tomatoes	Netflix	Hulu	Prime Video	Disney+	Type	Directors	Genres	Country	Language	Runtime
ID
1	Inception	2010	13+	8.8	87%	1	0	0	0	0	Christopher Nolan	Action,Adventure,Sci-Fi,Thriller	United States,United Kingdom	English,Japanese,French	148.0
2	The Matrix	1999	18+	8.7	87%	1	0	0	0	0	Lana Wachowski,Lilly Wachowski	Action,Sci-Fi	United States	English	136.0
3	Avengers: Infinity War	2018	13+	8.5	84%	1	0	0	0	0	Anthony Russo,Joe Russo	Action,Adventure,Sci-Fi	United States	English	149.0
4	Back to the Future	1985	7+	8.5	96%	1	0	0	0	0	Robert Zemeckis	Adventure,Comedy,Sci-Fi	United States	English	116.0
5	The Good, the Bad and the Ugly	1966	18+	8.8	97%	1	0	1	0	0	Sergio Leone	Western	Italy,Spain,West Germany	Italian	161.0

Make the column names uniform

Renaming the Columns

raw_movies_df.columns

Index(['Title', 'Year', 'Age', 'IMDb', 'Rotten Tomatoes', 'Netflix', 'Hulu',
       'Prime Video', 'Disney+', 'Type', 'Directors', 'Genres', 'Country',
       'Language', 'Runtime'],
      dtype='object')

raw_movies_df.columns = map(str.lower, raw_movies_df.columns)

Making all column names lower case for ease of use. Can also use: df.rename(str.lower, axis='columns')

raw_movies_df = raw_movies_df.rename(columns={'rotten tomatoes':'rotten_tomatoes','prime video': 'prime_video'})

raw_movies_df

	title	year	age	imdb	rotten_tomatoes	netflix	hulu	prime_video	disney+	type	directors	genres	country	language	runtime
ID
1	Inception	2010	13+	8.8	87%	1	0	0	0	0	Christopher Nolan	Action,Adventure,Sci-Fi,Thriller	United States,United Kingdom	English,Japanese,French	148.0
2	The Matrix	1999	18+	8.7	87%	1	0	0	0	0	Lana Wachowski,Lilly Wachowski	Action,Sci-Fi	United States	English	136.0
3	Avengers: Infinity War	2018	13+	8.5	84%	1	0	0	0	0	Anthony Russo,Joe Russo	Action,Adventure,Sci-Fi	United States	English	149.0
4	Back to the Future	1985	7+	8.5	96%	1	0	0	0	0	Robert Zemeckis	Adventure,Comedy,Sci-Fi	United States	English	116.0
5	The Good, the Bad and the Ugly	1966	18+	8.8	97%	1	0	1	0	0	Sergio Leone	Western	Italy,Spain,West Germany	Italian	161.0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
16740	The Ghosts of Buxley Hall	1980	NaN	6.2	NaN	0	0	0	1	0	Bruce Bilson	Comedy,Family,Fantasy,Horror	United States	English	120.0
16741	The Poof Point	2001	7+	4.7	NaN	0	0	0	1	0	Neal Israel	Comedy,Family,Sci-Fi	United States	English	90.0
16742	Sharks of Lost Island	2013	NaN	5.7	NaN	0	0	0	1	0	Neil Gelinas	Documentary	United States	English	NaN
16743	Man Among Cheetahs	2017	NaN	6.6	NaN	0	0	0	1	0	Richard Slater-Jones	Documentary	United States	English	NaN
16744	In Beaver Valley	1950	NaN	NaN	NaN	0	0	0	1	0	James Algar	Documentary,Short,Family	United States	English	32.0

16744 rows × 15 columns

Looking at a Snapshot of the Data

raw_movies_df.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 16744 entries, 1 to 16744
Data columns (total 15 columns):
 #   Column           Non-Null Count  Dtype  
---  ------           --------------  -----  
 0   title            16744 non-null  object 
 1   year             16744 non-null  int64  
 2   age              7354 non-null   object 
 3   imdb             16173 non-null  float64
 4   rotten_tomatoes  5158 non-null   object 
 5   netflix          16744 non-null  int64  
 6   hulu             16744 non-null  int64  
 7   prime_video      16744 non-null  int64  
 8   disney+          16744 non-null  int64  
 9   type             16744 non-null  int64  
 10  directors        16018 non-null  object 
 11  genres           16469 non-null  object 
 12  country          16309 non-null  object 
 13  language         16145 non-null  object 
 14  runtime          16152 non-null  float64
dtypes: float64(2), int64(6), object(7)
memory usage: 2.0+ MB

Cleaning the Rotten Tomatoes Column

raw_movies_df['rotten_tomatoes'] = raw_movies_df['rotten_tomatoes'].str.replace('%', '')

raw_movies_df.head()

	title	year	age	imdb	rotten_tomatoes	netflix	hulu	prime_video	disney+	type	directors	genres	country	language	runtime
ID
1	Inception	2010	13+	8.8	87	1	0	0	0	0	Christopher Nolan	Action,Adventure,Sci-Fi,Thriller	United States,United Kingdom	English,Japanese,French	148.0
2	The Matrix	1999	18+	8.7	87	1	0	0	0	0	Lana Wachowski,Lilly Wachowski	Action,Sci-Fi	United States	English	136.0
3	Avengers: Infinity War	2018	13+	8.5	84	1	0	0	0	0	Anthony Russo,Joe Russo	Action,Adventure,Sci-Fi	United States	English	149.0
4	Back to the Future	1985	7+	8.5	96	1	0	0	0	0	Robert Zemeckis	Adventure,Comedy,Sci-Fi	United States	English	116.0
5	The Good, the Bad and the Ugly	1966	18+	8.8	97	1	0	1	0	0	Sergio Leone	Western	Italy,Spain,West Germany	Italian	161.0

#converting to number, run as float first
raw_movies_df['rotten_tomatoes'] = raw_movies_df['rotten_tomatoes'].astype('float')
raw_movies_df['rotten_tomatoes'] = raw_movies_df['rotten_tomatoes'].astype('Int64')

raw_movies_df['rotten_tomatoes']

ID
1          87
2          87
3          84
4          96
5          97
         ... 
16740    <NA>
16741    <NA>
16742    <NA>
16743    <NA>
16744    <NA>
Name: rotten_tomatoes, Length: 16744, dtype: Int64

Now the Data is Clean, Let's Explore

movie_df = raw_movies_df.copy()

movie_df.describe()

	year	imdb	rotten_tomatoes	netflix	hulu	prime_video	disney+	type	runtime
count	16744.000000	16173.000000	5158.000000	16744.000000	16744.000000	16744.000000	16744.000000	16744.0	16152.000000
mean	2003.014035	5.902751	65.428461	0.212613	0.053930	0.737817	0.033684	0.0	93.413447
std	20.674321	1.347867	26.614496	0.409169	0.225886	0.439835	0.180419	0.0	28.219222
min	1902.000000	0.000000	2.000000	0.000000	0.000000	0.000000	0.000000	0.0	1.000000
25%	2000.000000	5.100000	44.000000	0.000000	0.000000	0.000000	0.000000	0.0	82.000000
50%	2012.000000	6.100000	71.000000	0.000000	0.000000	1.000000	0.000000	0.0	92.000000
75%	2016.000000	6.900000	88.000000	0.000000	0.000000	1.000000	0.000000	0.0	104.000000
max	2020.000000	9.300000	100.000000	1.000000	1.000000	1.000000	1.000000	0.0	1256.000000

How Much Data is Missing?

total = movie_df.isnull().sum().sort_values(ascending=False)
percent = (movie_df.isnull().sum()/movie_df.isnull().count()).sort_values(ascending=False)
missing_data = pd.concat([total, percent], axis=1, keys=['Total', 'Percent'])
missing_data

	Total	Percent
rotten_tomatoes	11586	0.691949
age	9390	0.560798
directors	726	0.043359
language	599	0.035774
runtime	592	0.035356
imdb	571	0.034102
country	435	0.025979
genres	275	0.016424
type	0	0.000000
disney+	0	0.000000
prime_video	0	0.000000
hulu	0	0.000000
netflix	0	0.000000
year	0	0.000000
title	0	0.000000

How Many Movies Are on Each Platform?

netflix_count = len(movie_df[movie_df['netflix'] == 1].index)
hulu_count = len(movie_df[movie_df['hulu'] == 1].index)
prime_count = len(movie_df[movie_df['prime_video'] == 1].index)
disney_count = len(movie_df[movie_df['disney+'] == 1].index)

label = ['Netflix', 'Hulu', 'Prime Video', 'Disney']
count = [netflix_count, hulu_count, prime_count, disney_count]
platform = pd.DataFrame(
    {'platform': label,
    'movie_count': count,
    })
platform

	platform	movie_count
0	Netflix	3560
1	Hulu	903
2	Prime Video	12354
3	Disney	564

fig = px.pie(platform, names='platform', values='movie_count')
fig.update_traces(rotation=45, pull=[0.1,0.03,0.03,0.03,0.03], textinfo='percent+label', title='Movie count per platform')
fig.show()

How Many Movies per Year?

yearly_count = movie_df.groupby('year')['title'].count().reset_index().rename(columns={'title':"movie_count"})
fig = px.bar(yearly_count, x='year', y="movie_count", color='movie_count', height=600)
fig.show()

What is the Average Rating Per Genre?

movie_df['genres'].value_counts().head(5)

Drama           1341
Documentary     1229
Comedy          1040
Comedy,Drama     446
Horror           436
Name: genres, dtype: int64

top5_genres = ['Drama','Documentary','Comedy', 'Comedy,Drama','Horror']
table = movie_df.loc[:, ['year', 'genres', 'imdb']]
table['avg_rating'] = table.groupby([table.genres, table.year])['imdb'].transform('mean')
table.drop('imdb', axis=1, inplace=True)
table = table[(table.year > 2009) & (table.year<2020)]
table = table.loc[table['genres'].isin(top5_genres)]
table = table.sort_values('year')

What about in the past decade?

fig = px.bar(table, x='genres', y='avg_rating', animation_frame='year', animation_group='genres', 
             color='genres', hover_name='genres', range_y=[0,10])
fig.update_layout(showlegend=False)
fig.show()

According to IMBd, What are the Best Movies?

top_movies_imdb = movie_df[movie_df['imdb']>8.5][['title', 'directors', 'rotten_tomatoes', 'imdb']].sort_values(by=['rotten_tomatoes'], ascending=False)
top_movies_imdb

	title	directors	rotten_tomatoes	imdb
ID
4474	Stop Making Sense	Jonathan Demme	100	8.6
4663	Tom Petty and the Heartbreakers: Runnin' Down ...	Peter Bogdanovich	100	8.6
3563	Parasite	Bong Joon Ho	99	8.6
5	The Good, the Bad and the Ugly	Sergio Leone	97	8.8
3562	GoodFellas	Martin Scorsese	96	8.7
...	...	...	...	...
11161	Draw Hard	Jon Nix	<NA>	8.7
11208	Along Recovery	Justin J. Springer	<NA>	8.6
11874	Ruby's Studio: the Feelings Show	Jason Docter,Matt Docter	<NA>	8.8
11911	Second Chance By Betrayal	Terrence Weasel Smith	<NA>	8.7
15176	Shred: The Story of Asher Bradshaw	Kathy Herndl	<NA>	8.8

97 rows × 4 columns

According to Rotten Tomatoes, What are the Best Movies?

top_movies_rt = movie_df[movie_df['rotten_tomatoes']>90.0][['title', 'directors', 'rotten_tomatoes', 'imdb']].sort_values(by=['rotten_tomatoes'], ascending=False)
top_movies_rt

	title	directors	rotten_tomatoes	imdb
ID
16481	Marvel Rising: Secret Warriors	Alfred Gimeno,Eric Radomski	100	5.2
1332	Attacking the Devil: Harold Evans and the Last...	David Morris,Jacqui Morris	100	8.1
5214	Top Spin	Sara Newens,Mina T. Son	100	6.9
5228	Bolivia	Israel Adrián Caetano	100	7.3
5229	The Story of Louis Pasteur	William Dieterle	100	7.3
...	...	...	...	...
3948	All Square	John Hyams	91	5.8
6111	Werewolf	Freddie Francis	91	6.0
6082	Michael H. – Profession: Director	Yves Montmayeur	91	7.0
5084	Cutter's Way	Ivan Passer	91	6.9
959	Nobody Speak: Trials of the Free Press	Brian Knappenberger	91	6.5

1035 rows × 4 columns

What Countries Produced the Most Movies?

movies_by_country = movie_df.groupby('country')['title'].count().reset_index().sort_values('title',ascending = False).head(10).rename(columns = {'title':'movie_count'})
fig = px.pie(movies_by_country,names='country', values='movie_count')
fig.update_traces(rotation=180, pull=[0.1,0.03,0.03,0.03,0.03],textinfo="percent+label", title='Movie Count per Country')
fig.update_layout(showlegend=False)
fig.show()

What is the Most Popular Movie Language in this Dataset?

count_by_language = movie_df.groupby('language')['title'].count().reset_index().sort_values('title', ascending=False).head(10).rename(columns={'title': 'movie_count'})
fig = px.bar(count_by_language, x='language', y='movie_count', color='movie_count', height=600)
fig.show()

What are the Longest Running Movies Represented?

top_runtime = movie_df.sort_values('runtime', ascending=False).head(10)
fig = px.bar(top_runtime, x='title', y='runtime', color='runtime', height=600)
fig.show()

Which Directors Have the Most Movies?

plt.figure(figsize=(15,7))
chains=movie_df['directors'].value_counts()[:20]
sns.barplot(x=chains, y=chains.index, palette='Set1')
plt.title('Most movies made by director', size=20, pad=20)
plt.xlabel('counts', size=15);